Google Search Appliance Connectors Administration Guide Google Search Appliance Connectors software version 4.0.3 Google Search Appliance software version 7.2

October 2014

Table of Contents About this Guide 1 About Connectors 4.0 The Lister/Retriever model 2 What’s New in Connectors 4.0? Resiliency Easy monitoring and reporting Off-Board installation Connector configuration Crawl configuration Automatic updates every 15 minutes Support for early binding only SAML security messages Simplified troubleshooting Giving all users access to all documents Saving feeds sent to the GSA 3 General Information Supported connectors Supported Google Search Appliance versions Configuration properties file Repository content relevancy Secure crawling and serving configurations Admin Console access Secure mode Supported ACL features Mark all documents as public Archive feeds Reverse proxy setup Download the connector software Run a connector as a service on Windows Register the connector as a service Add optional parameters Apache Daemon Logging Service Username and Password Jvm options Run the connector as a service

Stop running a connector 4 Enable Connector Security Certificate Authorities Self-signed certificates Create a self-signed certificate for the GSA Firefox Chrome OpenSSL (command line) Create a self-signed certificate for the connector Exchange certificates Turn on security with the server.secure property Run in secure mode with self-signed certificates Enable stricter security 5 Configure Connector Logs Logging properties file Change the location of logs Change the logging level Change the log file size Change the number of log files 6 Monitor Connectors with the Dashboard Supported browsers Dashboard port number Log in to the Connector Dashboard View information about a connector Version Status Statistics Connector configuration Recent log messages Start or restart feeds Encode sensitive values Download rich data about the connector 7 Troubleshoot Connectors Debug a connector by using a web browser Troubleshooting scenario Troubleshooting quick reference 8 Common configuration options

About this Guide This Administration Guide is intended for anyone who needs to understand how to manage Google Search Appliance (GSA) Connectors 4.0. It provides overview information about the Connectors, as well as procedures that you can follow to install, configure, or monitor each of the Connectors. The guide assumes that you are familiar with Windows or Linux operating systems and configuring the Google Search Appliance by using the Admin Console. For information about installing and configuring the connectors, see the following guides: ●

Deploying the Connector for SharePoint

●

Deploying the Connector for SharePoint User Profiles

●

Deploying the Connector for File Systems

●

Deploying the Connector for Active Directory

These guides, as well as information about using the Admin Console are available from the Google Search Appliance Help Center. For information about previous versions of connectors, see the Connector documentation page in the Google Search Appliance Help Center.

1 About Connectors 4.0 Google Search Appliance connectors enable the Google Search Appliance to acquire content from external repositories and provide that content in search results. A Google Search Appliance with configured connectors can perform fast, unified, secure search across multiple systems and document repositories. A fundamental strength of the search appliance is discovering enterprise content in web pages and indexing it. The GSA accomplishes this by crawling the web pages over HTTP/HTTPS, following hyperlinks within the pages to interrelated web pages, and adding the content it discovers to the search index. Ultimately, the GSA serves content from its index as search results to end users. However, many organizations have content that is stored in repositories, such as SharePoint and Windows file shares, rather than on web pages. Because documents in repositories are not usually interrelated through hyperlinks, the search appliance cannot find this content through normal crawling. Connectors 4.0 exploit the search appliance’s strengths by enabling it to crawl non-web content in repositories over HTTP/HTTPS. Additionally, connectors can feed groups information to the search appliance. Groups information can restrict the visibility of certain content to members of particular groups by using Access Control Lists (ACLs). The search appliance adds content acquired through connectors to the search index and uses credentials provided by connectors to protect secure content. There are several ways to model and communicate your repository's contents to the GSA, and Adaptors are one of them. For other possible solutions, look into Connectors 3.x and Content Feeds. Connectors 3.x support older GSA versions. Content Feeds should be used when the repository does not provide random document access and instead only provides changes occurring in the repository.

The Lister/Retriever model Connectors 4.0 are based on the lister/retriever model. In this model, the lister notifies the search appliance of the names of documents in the repository and the retriever sends URLs and ACLs in feeds to the search appliance. The search appliance uses the URLs to crawl documents in the repository over HTTP/HTTPS. Each document in a repository is identified by a unique document identifier (DocId). The search appliance performs the task of determining which DocIds the user can access and handles authentication requests between the search appliance and the repository. The following diagram shows how the connectors interact with the search appliance and the repository.

2 What’s New in Connectors 4.0? In release 4.0, connectors work seamlessly with more search appliance features than previous releases. Noteworthy features of Connectors 4.0 include: ●

Resiliency

●

Easy monitoring and reporting

●

Off-Board installation

●

Connector configuration

●

Crawl configuration

●

Updates every 15 minutes

●

Support for early binding only

●

SAML security messages

●

Simplified troubleshooting

●

Giving all users access to all documents

●

Saving feeds sent to the GSA

Resiliency Because Connectors 4.0 are “stateless,” they are more resilient than previous versions. For example, in previous versions, you had to restart the connectors when errors occurred. In this version, when an error occurs, you can ignore it because it will be corrected without administrator intervention.

Easy monitoring and reporting The Connector Dashboard is a new, web-based interface that eases monitoring and reporting. It displays the connector’s version, status, statistics, configuration variables and values, response time and available throughput in real time (including the previous hour and the previous day), and recent log messages. Also, from the Connector Dashboard, you can download the Diagnostics zip archive, which contains logs and other rich data about the connector. For more information about the Connector Dashboard, see Monitor Connectors with the Dashboard.

Off-Board installation All Connectors 4.0 are installed on a separate host server rather than the search appliance itself. For more details on this topic, see Download the connector software.

Connector configuration Because Connectors 4.0 are not built-in the search appliance, they are not configured through the search appliance Admin Console, as in previous releases. Configuration is handled in the adaptor-config.properties file. In release 4.0, “adaptor” is the internal name of the architecture used by the connectors.

Crawl configuration As part of deploying a connector, you configure crawl of repository content by using the search appliance Admin Console. Once crawling starts, you can use the search appliance’s robust features for monitoring and fine-tuning crawl. Take note that Connectors 4.0 crawl content according to the crawl schedule that is configured in the Admin Console. Because some repositories make it easy to notice what has changed recently, adaptors for those repositories can inform the GSA immediately when things happen, and even force the GSA to immediately recrawl a document because the adaptor knows that document was modified recently. This can allow the GSA to pick up changes to a repository with very low latency, but is not required for correct operation. For information about configuring crawl by using the Admin Console, see Administering Crawl.

Automatic updates every 15 minutes Every 15 minutes, the Connectors for SharePoint, File Systems, and Active Directory provide recent changes to the GSA. You can change the default time interval between updates by editing the adaptor.incrementalPollPeriodSecs configuration option.

Support for early binding only Connectors 4.0 enable support for authorization by early binding only; late binding is not supported. If your implementation requires late binding, Google recommends using a previous version of connectors.

SAML security messages In this release, SAML is the communication protocol between the search appliance and the connector for user authentication and authorization. This protocol replaces XML, which was used in previous releases.

Simplified troubleshooting A strength of Connectors 4.0 is simplified troubleshooting. If you encounter issues, you have several options for troubleshooting them, including the Connector Dashboard, as well as index and real-time diagnostics on the search appliance, and logs on the connector machine. For detailed information, see Troubleshoot Connectors.

Giving all users access to all documents Connectors 4.0.3 introduces the capability of giving all users access to all documents. For information about this capability, see Mark all documents as public.

Saving feeds sent to the GSA Connectors 4.0.3 introduces the capability of archiving all feeds sent to the GSA on the local drive. For more information, see Archive feeds.

3 General Information This section contains general information about Connectors 4.0, including: ●

Supported connectors

●

Supported Google Search Appliance versions

●

Configuration properties file

●

Repository content relevancy

●

Secure crawling and serving configurations

●

Admin Console access

●

Secure mode

●

Supported ACL features

●

Mark all documents as public

●

Archive feeds on the local drive

●

Reverse proxy setup

●

Download the connector software

●

Run a connector as a service

●

Stop running a connector

Supported connectors In version 4.0, Google supports the following connectors: ●

Connector for SharePoint

●

Connector for SharePoint User Profiles

●

Connector for File Systems

●

Connector for Active Directory

Guides for deploying each connector are available from the Google Search Appliance Help Center.

Supported Google Search Appliance versions Connectors 4.0.3 work with Google Search Appliance version 7.2.0.G.90 or higher. If you plan on deploying the Connector for Active Directory and you need to support over 1 million group memberships (which includes nested groups and users), then use GSA software version 7.2.0.G.230 or higher.

Configuration properties file Configuration is handled in the adaptor-config.properties file. Each connector installation procedure in this documentation contains steps for creating its adaptorconfig.properties file and adding minimal variables to make the connector operational. For example, after you create the configuration file, you need to add the server.port variable to point to the retriever port. For more information about configuration, see the section that pertains to your connector. Additionally, there are common configuration variables, which are used by all connectors. If you do not indicate values for these variables, defaults are used. For more information about this topic, see Common configuration options.

Repository content relevancy The search appliance determines the relevancy in search results of a document that it crawls on the web by using a "page rank" algorithm, which is based on an analysis of hyperlinks among documents. A search appliance administrator can view the relative page rank of a document by using the Index > Diagnostics > Index Diagnostics page in the Admin Console. By default, the search appliance sets the relevancy of a document that it crawls in a repository by using a "content rank" algorithm, which is based on analysis of the document's content. If a document has a content rank, its relative page rank on the Index > Diagnostics > Index Diagnostics page is zero. The configuration variable gsa.scoringType controls whether the search appliance uses a content rank algorithm or page rank algorithm for repository content. Valid values for this variable are: ●

content (content rank)

●

web (page rank)

If the connector uses the web scoring type, you must add the connector root URL as a Start URL on the Content Sources > Web Crawl > Start and Block URLs page in the GSA Admin Console. Otherwise, all documents will have a page rank of 1.

Secure crawling and serving configurations Connectors 4.0 support the authentication and authorization configurations for crawling and serving that the GSA administrator configures for the search appliance. For information about secure crawling and serving configurations, see Managing Search for Controlled-Access Content.

Admin Console access If the search appliance only allows HTTPS access to the Admin Console, then the connector must be running in secure mode. In secure mode, use HTTPS to access the Connector Dashboard. To disable HTTPS only access to the Admin Console, select Enable HTTP (i.e. non SSL) Admin Console and Version Manager access on the Administration > System Settings page in the Admin Console. When HTTP access to the Admin Console is enabled, you can use the Connector Dashboard with or without security enabled for the connector.

Secure mode Connectors 4.0 support communication in secure mode over HTTPS. You can enable secure mode for any connector, but Google strongly suggests that you enable security for the Connector for SharePoint, and the Connector for SharePoint User Profiles. For more information on this topic, see Enable connector security.

Supported ACL features Access Control Lists (ACLs) control which documents a user can see. The search appliance needs to crawl and index all documents, but still rapidly determine which documents a specific user is allowed to view in a search. All the 4.0 connectors feed ACLs at crawl time, using a separate channel from the other feeds. Access Control Lists (ACLs) may be inherited from a parent. This reduces the number of ACLs that require re-indexing. Connectors send full-fidelity ACLs, which include inheritance and can contain local groups. The connectors also support Deny ACLs - ACLs which deny access to specific individuals or groups, local and global namespaces for ACL users and groups and Active Directory groups.

ACLs are stored in the search appliance’s group database.

Mark all documents as public Adding the variable adaptor.markAllDocsAsPublic=true to the adaptorconfig.properties file enables you to treat all users as if they are members of all groups, thereby giving them access to all documents. The default value for adaptor.markAllDocsAsPublic is “false.” Take note that if this option is set to “true” for the Connector for Active Directory, the connector does not send a feed specifying group membership to the search appliance.

Archive feeds Adding the variable feed.archiveDirectory with a valid path to the adaptorconfig.properties file enables you to save feeds to the specified directory on the local drive as they are sent to the GSA. All feeds successfully and unsuccessfully sent to the GSA are archived. Failed feeds are tagged with FAILED in the archive feed file name. The feeds contain listed document-ids, named resources, and group definitions

Reverse proxy setup You can add a reverse proxy to your configuration as an intermediary for crawl requests from the search appliance to the connectors. For example, you might configure a proxy server (PROXYHOST) between the search appliance and multiple connectors in a roundrobin setup. To set up a reverse proxy: 1. Configure server.hostname as the proxy server instead of the server running

the connector. In the previous example, you would configure server.hostname=PROXYHOST 2. Optionally configure server.reverseProxyPort (defaults to server.port). This

option controls the port used in retriever URLs. 3. Optionally configure server.reverseProxyProtocol to either http or https,

depending on proxy traffic (defaults to https in secure mode and http otherwise).

Download the connector software All connectors 4.0 must be installed on a host machine. This connector version does not support installing connectors on the Google Search Appliance. To download the software for a connector, visit http://googlegsa.github.io/adaptor/index.html. Executables are available for all the 4.0 connectors. Google provides the installation software for each 4.0 connector in a single binary file, as listed in the following table. Repository

Connector

Executable

SharePoint

SharePoint

sp-install-4.0.3.exe

SharePoint User Profile Service Application

SharePoint User Profiles

spup-install-4.0.3

Microsoft Windows Shares

File System

fs-install-4.0.3.exe

Microsoft Active Active Directory Directory

ad-install-4.0.3.exe

For information about installing your connector, see the appropriate connector deployment guide, as listed in About this Guide.

Run a connector as a service on Windows When you run a connector as a service, you do not need to run it manually. The connector runs when the host server starts up, and shuts down when the host is shut down. Before running the connector as a service, register it, as described in the following section

Register the connector as a service You register the connector as a service by running the prunsrv command, as shown in the following procedure. Take note that you can increase the power of the prunsrv command by adding optional parameters. For example, you can specify logging for the Apache daemon by using the LogPath parameter. For detailed information , see “Add optional parameters.”

To register a connector as a service: 1. Download and extract prunsrv.exe from the latest Windows binary download of Apache Commons Daemon. If you are running on 64-bit Windows and will use a 64bit JVM, then you should use the prunsrv.exe in the amd64/ directory. 2. Place prunsrv.exe in the same directory as the connector you would like to run as a service. 3. In the same directory where the connector .jar files are installed, run the following command: prunsrv install ^ --StartPath="" ^ --Classpath= ^ --StartMode=jvm ^ --StartClass=com.google.enterprise.adaptor.Daemon ^ --StartMethod=serviceStart ^ --StartParams= ^ --StopMode=jvm ^ --StopClass=com.google.enterprise.adaptor.Daemon ^ --StopMethod=serviceStop ^ --StdOutput= ^ --StdError= ^ --Jvm= ^ --Startup=auto Where: is the name of the connector in the list of running services: SharePoint: adaptor-sharepoint SharePoint User Profiles: adaptor-sharepoint-user-profile File Systems: adaptor-fs Active Directory: adaptor-ad is the absolute path of the StartPath, for example "C:\Users\administrator.GSA\Desktop\Connector" is the the name of the connector .jar, for example, adaptorsharepoint-4.0.3-withlib.jar is one of the following values: com.google.enterprise.adaptor.sharepoint.SharePointAdaptor

com.google.enterprise.adaptor.sharepoint.SharePointUserProfileAda ptor com.google.enterprise.adaptor.fs.FsAdaptor com.google.enterprise.adaptor.ad.AdAdaptor is the full path to the output log, for example, C:\sp\logs\stdout.log is the full path to the error log, for example, C:\sp\logs\stderr.log is the path to where Java Virtual Machine dynamic link library is installed, for example, C:\Java\jdk1.7.0_67\jre\bin\server\jvm.dll An alternative to specifying the JVM on the command line with the Jvm parameter is to configure the default JVM with the Java Control panel (javacpl.exe). Be sure to update the service registration each time you update the JVM.

Add optional parameters You can add important optional parameters to the prunsrv command to specify: ●

Apache daemon logging

●

Service username and password

●

JVM options

Apache Daemon Logging You can specify logging for the Apache Daemon by using the LogPath parameter. The default value is: %SystemRoot%\System32\LogFiles\Apache where SystemRoot is a root path, for example C:\Windows Service Username and Password Use the ServiceUser parameter to specify the name of the account under which the service should run, as shown in the following example: --ServiceUser DOMAINname\username ^

Use the ServicePassword to specify the password for the account designated by the ServiceUser parameter, as shown in the following example: --ServicePassword password ^ Jvm options Use the JvmOptions parameter to specify a JvList of options in the form of -D or X that will be passed to the JVM, as shown in the following example: ++JvmOptions=-Djava.util.logging.config.file=logging.properties

Run the connector as a service To run a connector as a service, run the following command in the same directory where the connector .jar files are installed: prunsrv start

Stop running a connector To stop running a connector in Windows or Linux, close the connector command prompt on the host. You can stop running a connector as a service on Windows from either the service list or the command line. To stop running a connector as a service on Windows from the service list: 1. On the connector host, choose Start > Run > services.msc 2. Select the connector service. 3. Click Stop. To stop running a connector as a service on Windows from the command line, enter the following command on the host: prunsrv stop

Where is the internal name of the connector: ●

SharePoint: adaptor-sharepoint

●

SharePoint User Profiles: adaptor-sharepoint-user-profile

●

File Systems: adaptor-fs

●

Active Directory: adaptor-ad

To stop running a connector by using the search appliance Admin Console, perform either or both of the following actions: 1. On the Content Sources > Diagnostics > Crawl Status page, click Pause Crawl. 2. On the Content Sources > Web Crawl > Start and Block URLs page, remove the connector URL from the Follow Patterns.

4 Enable Connector Security In secure mode, the connectors communicate with the Google Search Appliance over HTTPS. You can enable security for any connector by configuring certificates and turning on security. Take note that you must enable security for the Connector for SharePoint and the Connector for SharePoint User Profiles. Secure mode supports using either of the following types of certificates: ●

Certificate Authorities (CA’s)

●

Self-signed certificates

In either case, you can also choose options to enable stricter security.

Certificate Authorities The GSA and the connector executable both have default Certificate Authorities; public keys are already in the GSA and connector trust stores. For the connector, you can find the default keystore CAs under jre\lib\security\. If you are using the default CA’s only, complete the tasks described in the following sections: ●

Exchange certificates

●

Turn on security

By default, the search appliance alias is “gsa” and the connector alias is “adaptor.” Optionally, you can configure either alias.

Self-signed certificates If you need to create self-signed certificates before turning on security, complete the tasks described in the following sections: ●

Create a self-signed certificate for the GSA

●

Create a self-signed certificate for the connector

●

Exchange certificates

●

Turn on security with the server.secure property

Create a self-signed certificate for the GSA For information about creating a self-signed certificate for the search appliance, see the GSA Admin Console help page for Administration > SSL Settings. To get the GSA's freshly-created certificate to add it as a trusted host for the connector, follow the procedure for your preferred browser or the command line.

Firefox 1. Navigate to the GSA's secure search: https://gsahostname/. A warning page appears with the following message: "This Connection is Untrusted." This message appears because the certificate is self-signed and not signed by a trusted Certificate Authority. 2. Click, "I Understand the Risks" and "Add Exception." 3. Wait until the "View..." button is clickable, then click it. 4. Change to the "Details" tab and click "Export...". 5. Save the certificate in your connector's directory with the name "gsa.crt". 6. Click Close and Cancel to close the windows.

Chrome 1. Navigate to the GSA's secure search: https://gsahostname/. A warning page appears with the following message: "The site's security certificate is not trusted!" In the location bar, there should be a padlock with a red 'x' on it. 2. Click the padlock and then click "Certificate Information." 3. Change to the "Details" tab and click "Export...". 4. Save the certificate in your adaptor's directory with the name "gsa.crt". 5. Click Close and Cancel to close the windows.

OpenSSL (command line) 1. Execute the following command: openssl s_client -connect gsahostname:443 < /dev/null 2. Copy the section that begins with -----BEGIN CERTIFICATE----- and ends with -----END CERTIFICATE----- (including the BEGIN and END CERTIFICATE portions) into a new file. 3. Save the file in your connector's directory with the name "gsa.crt".

Create a self-signed certificate for the connector Generate a self-signed certificate for the connector and export the newly created certificate. 1. Within the connector’s directory, run the following command: keytool -genkeypair -keystore keys.jks -storepass changeit keypass changeit -alias adaptor -keyalg RSA -validity 365 2. For "What is your first and last name?", enter the hostname of the connector’s computer. You are free to answer the other questions however you wish (including not answering them). 3. Answer "yes" to "Is CN=yourcomputershostname, OU=... correct?" 4. Still in connector’s directory, run the following command: keytool -exportcert -alias adaptor -keystore keys.jks -storepass changeit -keypass changeit -rfc -file adaptor.crt 5. Copy cacerts from Java to the connector's directory: For Windows, run the following command: copy PATH\TO\JRE\lib\security\cacerts cacerts.jks For Linux ,run the following command: cp PATH/TO/JRE/lib/security/cacerts cacerts.jks 6. To allow the connector to trust itself, run the following command: keytool -importcert -keystore cacerts.jks -storepass changeit file adaptor.crt -alias adaptor 7. When prompted Trust this certificate?, answer yes.

Exchange certificates To allow the connector to trust the search appliance: 1. On the connector host, run the following command: keytool -importcert -keystore cacerts.jks -storepass changeit file gsa.crt -alias gsa 2. When prompted Trust this certificate?, answer yes. To allow the search appliance to trust the connector: 1. In GSA Admin Console, click Administration > Certificate Authorities.

2. Under Add more Certificate Authorities, click Browse. 3. Navigate to the connector’s directory and select adaptor.crt. 4. Click Save.

Turn on security with the server.secure property You can turn on security for the connector by using server.secure property, which enables HTTPS and certificate checking. Add the following line to your adaptorconfig.properties file: server.secure=true When server.secure=true, the connector uses the GSA's authentication configuration and HTTPS for all communication. Also, when the value of server.secure is true, the following conditions apply: ●

You need to add the key to the connector keystore with an alias defined in the connector config file, server.keyAlias.

●

The connector runs on the configured port enforcing SSL.

●

The Connector Dashboard runs on the configured port enforcing SSL.

●

Feeds from the connector are forced to the search appliance secure Feedergate port (19902), even if the search appliance accepts feeds over HTTP.

●

The connector validates the search appliance’s certificate during the SSL handshake.

Run in secure mode with self-signed certificates If you are using one or more self-signed certificates in your configuration, you must run the connector with SSL settings, as shown in the following example command: (Windows): java ^ -Djava.util.logging.config.file=src/logging.properties ^ -Djavax.net.ssl.keyStore=keys.jks ^ -Djavax.net.ssl.keyStoreType=jks ^ -Djavax.net.ssl.keyStorePassword= ^ -Djavax.net.ssl.trustStore=.jks ^ -Djavax.net.ssl.trustStoreType=jks ^ -Djavax.net.ssl.trustStorePassword=changeit ^ -classpath adaptor-name-4.0.3-withlib.jar ^ com.google.enterprise.adaptor.name.NameAdaptor

(Linux / Unix systems): java \ -Djava.util.logging.config.file=src/logging.properties \ -Djavax.net.ssl.keyStore=keys.jks \ -Djavax.net.ssl.keyStoreType=jks \ -Djavax.net.ssl.keyStorePassword= \ -Djavax.net.ssl.trustStore=.jks \ -Djavax.net.ssl.trustStoreType=jks \ -Djavax.net.ssl.trustStorePassword=changeit \ -classpath adaptor-name-4.0.3-withlib.jar \ com.google.enterprise.adaptor.name.NameAdaptor

Enable stricter security Optionally, you can improve security by choosing stricter security features on the Administration > SSL Settings page in the Admin Console, as described in the following table. However, using any of these options require the connector to be configured for security and have server.secure=true in its configuration.

Option

Setting

Description

Enable HTTP (non-SSL) access for Feedergate

Uncheck

When this option is unchecked, only HTTPS communications will be accepted by feedergate. Connectors send document ids to feedergate.

Enable Client Certificate Authentication for Feedergate

Check

When this option is checked, the Feedergate SSL port (19902) only accepts connections from IP addresses in the trusted IP addresses list and clients who present a valid x509 certificate when connecting. Valid means that the certificate is signed by a certificate in the CA keystore on the search appliance (or a certificate in the certificate chain).

Enable Server Certificate Authentication

Check

When this option is checked, it is a requirement for the crawler to authenticate certificates presented by servers that contain secure content.

You must include server.secure=true in the connector configuration before enabling these stricter features. To enable stricter security, perform the following steps by using the GSA Admin Console: 1. Click Administration > SSL Settings. 2. Make any of the following changes on this page: a. Uncheck Enable HTTP (non-SSL) access for Feedergate. b. Check Enable Client Certificate Authentication for Feedergate. c. Check Enable Server Certificate Authentication. 3. Click Save.

5 Configure Connector Logs The connectors log processing messages, including exceptions and warnings. Log messages appear in the Connector Dashboard and you can download the logs, as described in Download rich data about the connector. Messages contain information about thread processing, including: ●

Date stamp--Date and time the message was logged

●

Name-of-thread--The thread that generated the message

●

Last-30-characters-of-method--Code source for connector request

●

Logging level--Filter log messages by level of severity

●

Log-message--Text message for log entry

The following example shows a log message: 06-12 18:20:08.839 background URLConnection.getInputStream() FINE: sun.net.www.MessageHeader@b20ccbf12 pairs:...

Logging properties file Log configuration is controlled by the logging.properties file. Each connector installation procedure in this documentation contains a step for editing logging.properties. By editing values in this file, you can configure the following settings: ●

Location of logs

●

Logging level

●

Log file size

●

Number of log files

The following example shows a logging.properties file with default values. .level=INFO handlers=java.util.logging.FileHandler,java.util.logging.ConsoleHandler java.util.logging.FileHandler.formatter=com.google.enterprise.adaptor.CustomF ormatter java.util.logging.FileHandler.pattern=logs/adaptor.%g.log java.util.logging.FileHandler.limit=10485760

java.util.logging.FileHandler.count=20 java.util.logging.ConsoleHandler.formatter=com.google.enterprise.adaptor.Cust omFormatter com.google.enterprise.adaptor.CustomFormatter.useColor=true

Change the location of logs By default, the logs are saved in logs/adaptor.*.log, in the same directory where the connector is running. To change the location of log files, edit the java.util.logging.FileHandler.pattern value in the logging.properties file: java.util.logging.FileHandler.pattern=logs/adaptor.%g.log

Change the logging level You can filter messages written to log files by the following Java log levels: ●

FATAL

●

WARNING

●

INFO

●

FINE

●

FINER

●

FINEST

By default, the log level is INFO. The number of messages generated increases with each level, where FATAL logs the smallest number of messages and FINEST logs the largest. To change the level of log files, edit the .level value in the logging.properties file: .level=INFO

Change the log file size By default, the size of connector log files is 10485760 bytes. Restarting the connector will create a new log file, regardless of how large the previous one had been. You can change the size to suit your needs. The limit must be specified as a 32-bit integer, and thus has an upper limit of 2,147,483,647 (2 gigabytes, about 205 times as large as the default size).

To change the size of log files, edit the java.util.logging.FileHandler.limit value in the logging.properties file: java.util.logging.FileHandler.limit=10485760

Change the number of log files The connector writes to a log file until the size limit is reached, then starts writing to a new log file. By default, the connector writes to 20 log files, but you can change the number to suit your needs. There is no upper limit to the number of log files. After it finishes writing to the last log file, it starts writing over the first file. To change the number of log files, edit the java.util.logging.FileHandler.count value in the logging.properties file: java.util.logging.FileHandler.count=20

6 Monitor Connectors with the Dashboard The Dashboard is a web-based interface that provides information about the connector’s operation, with easy access to logs and error history. Use the Connector Dashboard to perform the following tasks: ●

View information about the connector

●

Start or restart feeds

●

Encode sensitive values

●

Download rich data about the connector

You must start the connector to use the Dashboard.

Supported browsers The Connector Dashboard runs in the following browsers: ●

Google Chrome 22

●

Internet Explorer 8 and 9

●

Firefox 15 and 16

●

Safari 5 and 6

Dashboard port number By default, the Connector Dashboard uses port 5679. The port number is determined by the value of the variable server.dashboardPort in the adaptor-config.properties file for the connector. You can change the Connector Dashboard port number by changing the default value in this file. Every instance of a connector running on a host must have a unique value for server.dashboardPort.

Log in to the Connector Dashboard To display the Connector Dashboard, open a browser and navigate to the following HTTP or HTTPS address: http://:/dashboard or https://:/dashboard

where: ●

HTTP or HTTPS--If you run the connector in secure mode, use HTTPS to log in to the Dashboard.

●

is the hostname or IP address of the host that is running the connector

●

is the dashboard port number, as specified in the adaptorconfig.properties file for the connector

To log in to the Connector Dashboard, use your search appliance user or administrator login credentials. You cannot log in to the Connector Dashboard with search appliance manager login credentials.

View information about a connector You can use the Dashboard to monitor the connector by viewing up-to-date information, including: ●

Version

●

Status

●

Statistics

●

Connector (Adaptor) configuration

●

Recent log messages

Version In the Version section, the Dashboard displays information about the currently installed Java version, Connector (“Adaptor”) library version, Connector type, and Connector version.

Status In the Status section, the Connector Dashboard displays the current status of the Java version (supported or not), feed pushing, the error rate of document retrieval from the repository (derived from logs), and search appliance crawling.

For each item, a signal indicates the status by color: ●

Green for OK. The item is functioning.

●

Yellow for alert. The item is not currently functioning, but no action is required. For example, the Dashboard displays yellow when the GSA is not currently crawling.

●

Red for warning. The item is not functioning and requires attention.

Statistics In the Statistics section, the Connector Dashboard displays the following information: ●

A datestamp for when the connector program was started.

●

Datestamps for the last successful push (full or incremental) start or end. The push can either be started automatically or manually.

●

Status of the current push, if any.

●

Total number of DocIDs pushed from the repository to the connector since the program started.

●

Total number of requests for documents and unique documents from the GSA and the connector.

●

Time resolution

The Statistics section also displays graphs showing throughput and response time for the last minute, last hour, and last day.

Connector configuration In the configuration section, the Connector Dashboard displays the values for all the configuration variables in the adaptor-config.properties file.

Recent log messages In the Recent Log Messages section, the Dashboard displays connector log messages. For more information on this topic, see Download rich data about the connector.

Start or restart feeds The Connector Dashboard enables you to start or restart a full feed or an incremental feed as often as needed or when errors are detected. To start or restart a feed, click either Run Incremental Push or Run Full Push.

Encode sensitive values You can encode passwords and other sensitive configuration values and copy them to the adaptor-config.properties file. Values can be specified in the configuration as prefix:data, where the prefix specifies how the value is stored. You can encode the listed sensitive values for the following connectors: ●

Connector for SharePoint--sharepoint.password

●

Connector for SharePoint User Profiles--sharepoint.password

●

Connector for Active Directory--ad.defaultPassword

●

The Connector for File Systems does not support encoding sensitive values.

The value can be stored as: ●

Plain text allowing the password or other information to be read by anybody who can read the configuration. Denoted by "pl" prefix.

●

Obfuscated where the information is in a highly unreadable format, but it is possible for anyone to retrieve the original text. Denoted by "obf" prefix.

●

Encrypted which uses your HTTPS encryption key to encrypt the value. Denoted by "pkc" prefix.

To encode a sensitive value: 1. Under Storing Sensitive Values, enter the sensitive value in the field. 2. Click a storage option. 3. Click Encode Sensitive Value. The encoded value appears. 4. Copy and paste the sensitive value into the adaptor-config.properties file.

Download rich data about the connector The Diagnostics zip archive contains rich data about the connector, including: ●

Current configuration settings (in the config.txt file)

●

Connector version, status, and statistics (in the state.txt file)

●

Thread details (in the threaddump.txt file)

●

Logs folder

This data that can help you to diagnose connector issues. To download the archive, click Diagnostics zip file on the Dashboard.

7 Troubleshoot Connectors Connectors 4.0 provide several options for troubleshooting issues, including: ●

Connector Dashboard for checking the status of feeds and document retrieval

●

Logs on connector machine for checking messages about thread processing

●

Search appliance index diagnostics for checking crawl status

●

Search appliance real-time diagnostics for checking HTTP headers for a specific URL at any time without having to wait for the crawler to ingest it

●

Web browser with the connector host

Additionally, you can troubleshoot issues by examining URL-and-metadata feed files. Because these types of feed files are relatively small, troubleshooting them does not require significant effort.

Debug a connector by using a web browser A connector, by default, will deny all document accesses, except from the search appliance. To allow debugging and testing a connector by using a browser without a search appliance, you can add a hostname to the server.fullAccessHosts configuration option to allow that computer full access to all connector content. In addition, this setting allows that computer to see metadata and other GSA-specific information as HTTP headers. This capability can be very useful when combined with Firebug or the Web Inspector in your browser to observe a connector's behavior.

Troubleshooting scenario In this scenario, users cannot find a specific document in search results, even though it is assumed to be in the search appliance index. To troubleshoot this issue, the administrator can track the document through the system by following the path a document takes to get into the search appliance index. The administrator might perform one or more of the following steps: 1. Make sure that the search appliance is set to follow and crawl the Connector's URLs by checking the Content Sources > Web Crawl > Start and Block URLs page in the Admin Console.

2. Make sure GSA crawling is not paused by using the Content Sources > Diagnostics > Crawl Status page. 3. Check the Connector status and recent log messages by using the Dashboard. 4. Ensure that the Connector fed the document URL to the search appliance by examining the feed file. 5. Ensure that the search appliance got the document by using the Index> Diagnostics > Index Diagnostics page in the Admin Console. 6. Check the HTTP header for the document by using the Content Sources > Diagnostics > Real-time Diagnostics page in the Admin Console. 7. Ensure that the connector logged the document by checking connector log files. The Lister logs a file when it feeds it to the search appliance. The Retriever logs the file when the crawler requests the document. 8. Find out if the connector has information about the document by using a web browser to access the file information on the connector host. If the document isn’t located, the administrator can request a recrawl of the missing document by restarting the crawl from the Connector Dashboard, or recrawling the URL by using the Content Sources > Web Crawl > Freshness Tuning page in the Admin Console.

Troubleshooting quick reference Error message/Issue

Resolution

Type of Connector

Logs: Unathorized request. Status code:200

Add host IP to the GSA's feeds' list of trusted IP addresses by using the Content Sources > Feeds page in the Admin Console.

SharePoint, SharePoint User Profiles, File Systems

Index diagnostics: Error: Permanent DNS failure.

Add a DNS override by using the Administration > DNS Override page in the Admin Console.

SharePoint

Index diagnostics: Retrying URL: Connection reset by peer during fetch.

DNS override is wrong. Correct it by using the Administration > DNS Override page in the Admin Console.

SharePoint

Errors in the logs for some

Host load is too high, try to

SharePoint

documents:The server sent HTTP reduce host load status code 503: Service unavailable Feeds are not coming through

●

Make sure GSA can accept feeds from the connector host machine.

●

Check connector logs for errors, such as failure to

SharePoint, SharePoint User Profiles, File Systems, Active Directory

connect to look-up GSA, or failure to communicate with the repository. Documents are not getting indexed

●

Make sure GSA is set to follow and crawl the Connector's URLs by checking the Content Sources > Web Crawl > Start and Block URLs page in the Admin Console.

●

Make sure GSA crawling is not paused by using the Content Sources > Diagnostics > Crawl Status page.

●

Check for error messages on the Index> Diagnostics > Index Diagnostics page.

●

Take a look at connector's log messages.

●

Check the Content Sources > Diagnostics > Real-time Diagnostics page for the particular URL that you expect to be indexed.

SharePoint, SharePoint User Profiles, File Systems

Crawling is slow

SharePoint, File Systems

Use the Dashboard to find: ●

What is the mean duration of a request (Response Time)? A couple hundred milliseconds would be good.

●

What is the max duration of a request? A file taking over a couple of minutes would be bad.

Document retrieval times out

The connector gives a document retrieval request 30 seconds to

SharePoint, File Systems

start and 3 minutes to complete. If you want to give your repository more time you can adjust adaptor.docContentTimeoutSecs

and adaptor.docHeaderTimeoutSecs

.

The Google Search Appliance

Files and folders that are marked

Index Diagnostics shows many

as hidden are not fed to the GSA.

documents with a Crawl Status

However, they may be listed on

of "Document not found (404)."

the Index> Diagnostics > Index

SharePoint, File Systems

Diagnostics page with a crawl status of "Document not found (404)." SharePoint is returning 401

Ensure that the full read

(unatuhorized)

permissions are given on the

SharePoint

SharePoint Web Application policy. Renamed user names are not

Run User Profile Synchronization

reflected in ACLs.

job for incremental updates.

SharePoint

8 Common configuration options The following table lists common configuration options, which are used by all connectors. If the administrator doesn’t set these options, defaults are used. The only required option is gsa.hostname. All others are optional. Name

Meaning

Default

gsa.acceptsDocControlsHeader

Use X-Gsa-Doc-Controls HTTP

true

header with namespaced ACLs. Otherwise ACLs are sent without namespace and as metadata. If not set, then an attempt to compute from gsa.version is made. adaptor.fullListingSchedule

When to invoke

03***

Adaptor.getDocIds, in cron format (minute, hour, day of month, month, day of week). adaptor.incrementalPoll PeriodSecs

Number of seconds between

900

invocations of PollingIncrementalLister.get ModifiedDocIds.

adaptor.docContentTimeoutSecs

Number of seconds a

180

connector has to complete sending content before it is interrupted. Timing starts when sending content starts. adaptor.docHeaderTimeoutSecs

Number of seconds connector

30

has to start sending content before it is interrupted. adaptor.markAllDocsAsPublic

When the value is “true,” all

false

documents are marked as “public.” Take note that if this option is set to “true” for the Connector for Active Directory, the connector does not send users/groups to specify group memberships. adaptor.pushDocIdsOnStartup

Whether to invoke

true

Adaptor.getDocIds on process start (in addition to adaptor.fullListingSchedul e). docId.isUrl

If your connector document ids false are already URLs, prevent them from being inserted into connector generated URLs.

feed.archiveDirectory

Save feeds of listed documentids, named resources, and group definitions to the specified directory on the local drive as they are sent to the GSA. All feeds successfully and unsuccessfully sent to the GSA are archived. Failed feeds are tagged with FAILED in the archive feed file name.

feed.crawlImmediately BitEnabled

Send bit telling GSA to crawl

feed.maxUrls

Set the maximum number of

immediately.

URLs included per feed file. feed.name

false

Source name used in feeds. Generated if not provided.

5000

feed.noRecrawlBitEnabled

Send bit telling the GSA to

false

crawl your documents only once. gsa.version

Version number used to

Defaults to

configure expected GSA

acquiring from

features.

GSA. Uses 7.0.14114 if acquiring fails.

gsa.characterEncoding

Character set used in feed files. UTF-8

gsa.hostname

Machine to send feed files to. Process errors if not provided.

gsa.samlEntityId

The SAML Entity ID that

http://google.com/

identifies the GSA.

enterprise/gsa/ security-manager

journal.reducedMem

Avoid tracking per URL

true

information in RAM; suggested with over five hundred thousand documents. gsa.scoringType

Type of relevance algorithm

content

GSA utilizes to rank documents. Either content or web. Is sent when adaptor.sendDocControlHeade r is true. server.dashboardPort

Port on a connector 's machine for accessing a connector's dashboard. Every instance of a connector running on a machine must have a unique value for server.dashboardPort.

5679

server.docIdPath

Part of URL preceding encoded

/doc/

document ids. server.fullAccessHosts

server.hostname

server.keyAlias

Hosts allowed access without

empty, but

authentication (certificates still

implicitly contains

needed when in secure mode).

gsa.hostname

Hostname of a connector

lowercase of

machine for URL generation.

automatically

The GSA will use this hostname

detected

to crawl the connector.

hostname

Keystore alias where

connector

encryption (public and private) keys are stored. server.maxWorkerThreads

Number of maximum

16

simultaneous retrievals allowed. server.port

Retriever port. Every instance

5678

of a connector running on a machine must have a unique value for server.port. server.queueCapacity

Maximum retriever queue size.

160

server.reverseProxyPort

Port used in retriever URLs (in

server.port

case requests are routed through a reverse proxy). server.reverseProxyProtocol

Can be either http or https,

http in secure

depending on proxy traffic.

mode or http otherwise

server.samlEntityId

server.secure

The SAML Entity ID that the

http://google.com/

connector uses to identity

enterprise/gsa/

itself.

adaptor

Enables https and certificate

false

checking. server.useCompression

Compress retrieval responses.

true

transform.acl.X

Where X is an integer, match

no modifications

and modify principals as described. transform.pipeline

Sequence of transformation

empty string (no

steps.

pipeline)